DE102008015286A1 - Optoelectronic sensor - Google Patents

Abstract

The sensor has two light transmitters (12) for sending light signals into a monitoring region. Codes of different transmitters are generated by quasi-orthogonal Hadamard-Walsh-codes or one of the codes in form a pseudo random code is assigned to the respective transmitter, where the pseudo random code forms a pattern sequence, which is divided into code segments. The other code is assigned to the other transmitter, and is generated by cyclic movement of the segments. The codes of the transmitters form a quasi-orthogonal code.

Description

The The invention relates to an optoelectronic sensor with several Light transmitters for emitting light signals in a surveillance area, at least one light receiver for receiving from the Lichtsendern emitted light signals and a control and / or Evaluation device for generating and / or influencing the light signals and for generating an object detection signal in case of Presence of an object in the surveillance area in Dependence on the output signals of the light receiver.

such In particular, sensors can be used as multi-beam light grid systems be used for access control, in particular dangerous machine work areas or specific spaces within of a building.

If in sensors of the type mentioned an object in the surveillance area and thereby the light path between light transmitter and light receiver is interrupted, an object detection signal is generated, For example, to shut down a machine and / or to trigger an audible or visual warning signal may result.

In the currently used light grids, a light pulse or a light pulse sequence is emitted which is or are detected by means of a threshold value detector. However, the respective light pulses can now be superimposed by interference signals, which makes the detection of the received pulses or pulse sequences at least more difficult. In the DE 199 26 214 A1 For the suppression of interference signals it is proposed to use so-called well-correlated chirp sequences as transmission signals. The useful signals are subjected to a spectral spread in order to make them insensitive to narrowband or impulse noise.

Out The mobile technology is the application of so-called spread spectrum techniques known. In this case, for example, a pseudo-random noise code over be generated a so-called primitive polynomial, with such a Pseudo-random noise code may include, for example, an M sequence.

It it is now conceivable, for example, to generate different M sequences, to produce distinguishable orthogonal transmission signals and to ensure that adjacent light grids do not interfere. Be there for the light grids from the field of mobile technology known spread spectrum techniques using a so-called CDMA (Code Division Multiple Access) method, in principle Kasami sequences, gold codes, M sequences, Barker codes, a Hadamart-Walsh code or a combination of these Codes are used as pulse sequences.

The orthogonality increases with increasing length of the pulse sequences. However, with light curtains nowadays, a fast response time is important, so it is not possible to use pulse sequences of any length. Rather, in practice, the pulse sequences are limited to a length of, for example, 2 ^N -1 chips, where N may be in the range from 5 to 8, for example. There is therefore the danger that the individual rays of a light grid interfere.

Of the Invention is based on the object, an improved optoelectronic To provide sensor of the type mentioned, as far as possible simple structure and the largest possible range an optimal signal to noise ratio of the individual beams of a Light grid ensured.

According to the invention This object is achieved in that at least two light emitters are provided, at least substantially simultaneously in the surveillance area emitted light signals each by one after the spread spectrum method generated with a code modulated output, wherein the Codes of different light transmitters by quasi-orthogonal Hadamard-Walsh codes are generated or a first light transmitter, a first code in the form one comprising an M sequence, gold sequence or Kasami sequence Pseudo-random codes are assigned, one in several code segments divided mother sequence forms, and at least one other Light emitter at least differing from the first code another code is assigned by a cyclic shift the code segments of the mother sequence is generated, the the different light sources associated with different codes form again quasi-orthogonal codes.

The at least substantially simultaneously emitted codes are thus selected in a particular simple manner and are evaluated so that they exert at least substantially no influence on each other. The sequences or sequences forming the codes can also be relative to one another during a shift still be clearly identified. Thus, it is now also achieved, above all, that the individual beams of a relevant light grid no longer interfere at least essentially, ie, the functionality of the sensor is essentially ensured.

The Control and / or evaluation, for example, so executed be that with n at least substantially simultaneously transmitted codes On the receiving side n different correlation spectra are evaluated.

there is the control and / or evaluation advantageously so executed that the n different correlation spectra over linked the positions of the individual correlation peaks be evaluated.

From Advantage is especially when the control and / or evaluation is designed so that at least substantially simultaneously emitted cyclically shifted M sequences, gold sequences or Kasami sequences on the receiving side all codes with only an evaluation process are identifiable. The evaluation can ie in particular by means of a single correlation process only a correlation spectrum can be performed. It In particular, it should be ensured that the receiving side all codes recorded with only one sensor and / or be sampled and linked to the evaluations of the individual codes can be.

at simultaneously emitted cyclically shifted M sequences, gold sequences or Kasami sequences can be the quasi-orthogonal property of different codes, especially on the autocorrelation properties M sequences, Gold sequences or Kasami sequences are based.

Prefers is the mother sequence in at least two, especially in at least three and preferably in two or three to six code segments divided up.

By the cyclic shift of the code segments of the mother sequence are advantageously at least two, in particular three and preferably generates two or three to six different codes.

The different codes are preferred different light emitters assigned to a light grid.

According to one preferred practical embodiment of the invention Optoelectronic sensor are those of the light emitters at least essentially simultaneously sent to the surveillance area Light signals in each case by one after the DSSS (Direct Sequence Spread Spectrum) band spreading method with the relevant code modulated Output signal generated.

According to this DSSS spread spectrum method can use a respective output signal a given bit sequence are spread. This bit sequence becomes also referred to as spreading code or chipping sequence.

The from the light emitters simultaneously into the surveillance area emitted light signals are expediently each by at least one mixed with the code in question Generated light pulse. In this case, the respective output signal So formed by at least one light pulse.

Conveniently, in particular the sum of the signal from the light receiver as sum signal received light signals to identify the one Light transmitter associated output signal with the relevant Light transmitter associated code applied.

there the control and / or evaluation device preferably comprises means to identify the associated with a respective light emitter Output signal the autocorrelation function between the received To form light signals or sum signal and the code in question.

Conveniently, is at least two light emitters and preferably each at least three light transmitters a common light receiver assigned.

at at least substantially simultaneously emitted quasi-orthogonal Hadamard Walsh codes, the synchronization of the sensor or Light grid expediently over individual Pseudo-random sequences.

at at least substantially simultaneously emitted cyclically shifted M-sequences, gold sequences or Kasami sequences are synchronized of the sensor or light grid preferably over that of the associated common light receiver received sum signal.

light source and photoreceptors may be at least partially on opposite sides of the surveillance area be arranged. In principle, it is also possible Light emitter and light receiver on one side of the surveillance area and a retroreflector on the opposite side of the surveillance area to arrange. The second variant corresponds to an autocollimation arrangement.

As already mentioned, are the light emitter and light receiver preferably provided in a light grid arrangement.

Prefers each light transmitter is a common light receiver assigned and according to the number of light receivers less than the number of light emitters.

So if so-called spread spectrum techniques in light grids are used as CDMA methods, then in principle Kasami sequences, gold codes, M sequences, Barker codes, a Hadamard-Walsh code or a combination of these codes can be used as pulse sequences become. However, as light curtains require a fast response time, it is not possible to use arbitrarily long pulse sequences. The pulse sequences are therefore limited in practice to the length of 2 ^N - 1 chips (N = 5, 6, 7, 8 (max)). This means that the pulse sequences can only be quasi-orthogonal. Now, a receiver of several, for example, 3 to 6 transmitters are to be irradiated, whereby the individual signals should be clearly identifiable by signal evaluation in the sense of the CDMA method.

The Orthogonality is to be considered in two ways. So the codes and the evaluation as well as the timinig be chosen so that several systems or light barriers, which are operated spatially side by side, no or have little influence on each other. In order to the systems are orthogonal to each other, that is it is insofar as the orthogonality outside of a respective Systems guaranteed.

To the another is said to have orthogonality within each one Systems, that is, for example, within a light grid be given. At the same time, the simultaneously transmitted codes should be essentially no influence on each other so that the consequences are clearly identified even with a relative shift can be. Here are the sequences or sequences to each other orthogonal.

There only quasi-orthogonal codes can be generated and used The system may be affected by the simultaneous transmission of signals even disturbing, which is particularly critical, since the fluctuations of the Light energy of the LEDs used due to the manufacturing process subject to natural fluctuation margins. So, for example It can be assumed that a maximum noise level of around a factor 2 stronger than the actual transmitter in the Receiver injects, which means that the user code by one Factor may be weaker than the sturgeon code. This may cause the recipient from time to time Fault code recognized as a useful signal. The application of the DSSS (Direct Sequence Spread Spectrum) / CDMA (Code Division Multiple Access) method leads to a minimization of the FAR (False Acceptance Rate).

by virtue of the solution according to the invention becomes a Reduction of susceptibility due to simultaneous transmission Codes reached, so that in light curtains an application of the DSSS (Direct Sequence Spread Spectrum) band spreading method as CDMA (Code Division Multiple Access) method is possible.

• – Es werden quasi-orthogonale Hadamard-Walsh-Codes im synchronisierten Betrieb verwendet, wobei zur Synchronisation einzelne pn- oder Peusdozufalls-Sequenzen herangezogen werden können.
• – Es werden quasi-orthogonale verschobene M-Sequenzen im synchronisierten Betrieb verwendet, wobei zur Synchronisation das Summensignal aller simultan gesendeten Codes herangezogen werden kann.
• – Bei n simultan gesendeten Codes werden n verschiedene Korrelations-Spektren ausgewertet. Da die Codes auch simultan empfangen werden, muss die Position der einzelnen Peaks an gleicher Position sein. Durch die Positionen können die unterschiedlichen Spektren verknüpft ausgewertet werden, was die FAR (False Acceptance Rate) deutlich minimiert.

In this case, according to the invention, in particular the following strategies can be used to preclude the recognition of a sturgeon code as a useful code:

• - Quasi-orthogonal Hadamard-Walsh codes are used in synchronized operation, with individual pn or Peusdo random sequences can be used for synchronization.
• - Quasi-orthogonal shifted M-sequences are used in synchronized operation, wherein for synchronization the sum signal of all simultaneously transmitted codes can be used.
• - For n simultaneously transmitted codes n different correlation spectra are evaluated. Since the codes are also received simultaneously, the position of the individual peaks must be in the same position. Due to the positions, the different spectra can be evaluated linked, which significantly reduces the FAR (False Acceptance Rate).

In shifted M-sequences, peaks of all the signals are sent simultaneously in a correlation spectrum codes, so that all codes can be identified with only one evaluation process, whereby here a constant distance of the individual peaks is decisive. The evaluation can thus be carried out in particular by means of a single correlation process with only one correlation spectrum.

Due to the solution according to the invention, a number of advantages are achieved. Thus, the transmission time in the light channel is reduced by the simultaneously transmitted signals. In addition, the simultaneous transmission of signals results in a reduction in the number of expensive reception ASICs. It has been shown that even with signals transmitted simultaneously, for example with 30 mm optics, the range is increased from 20 m to more than 100 m. For example, at N = 7 (N = ordering of the generator polynomial), up to 16 codes are possible for use of the DSSS method, whereby each of these codes can generate by cyclic shift up to six further shifted M-sequences. For example, up to six codes or shifted M-sequences can be sent simultaneously. By linking the correlation spectra results a very low FAR (False Acceptance Rate) even when a jets of jammers IEC 61496 ,

The Invention will be described below with reference to exemplary embodiments explained in more detail with reference to the drawing; in this show:

1 1 is a signal flow diagram of an embodiment of a system comprising a transmitter unit with three light transmitters, an optical channel and a light receiver unit with a light receiver;

2 an example of a binary code,

3 an example of a code signal,

4 an example of a sum signal,

5 an example of a discrete sum signal,

6 an example of a reference code,

7 a signal flow diagram of an exemplary evaluation,

8th a correlation spectrum for the definition of terms,

9 a state graph of the system,

10 a representation of a code signal and reference code with a shift τ,

11 a correlation spectrum with a shift τ of the peak position,

12 an exemplary course of the peak position under ideal conditions,

13 an exemplary course of the peak position with different disturbances,

14 the principle of shifted M sequences,

15 an exemplary course of the peak position with synchronization to a code signal,

16 an exemplary course of the peak position with synchronization to the sum signal,

17 exemplary correlation spectra of the three codes,

18 an extended computation of the PSR for shifted M-sequences,

19 the principle of occurring interference peaks, and

20 a FAR-FRR map of shifted M-sequences with linked evaluation.

1 shows a signal flow diagram of an embodiment of a transmission unit 10 with three optotransmitters 12 , an optical channel 14 and a light receiving unit 16 with a light receiver 18 comprehensive system. Such a system is part of an optoelectronic sensor that may include one or more such systems. In principle, a respective system may also have only two or more than three light transmitters 12 include. In addition, the optoelectronic sensor with a control and / or evaluation 20 be provided. The optoelectronic sensor may in particular be a light grid.

A relevant optoelectronic sensor can therefore in particular have a plurality of light transmitters 12 for emitting light signals in a surveillance area, at least one light receiver 18 to receive from the light emitters 12 emitted light signals and a control and / or evaluation 20 for generating and / or influencing the light signals and for generating an object detection signal in the case of the presence of an object in the monitoring area in dependence on the output signals of the light receiver 18 include.

In this case, according to the invention at least two light emitters 12 provided, at least substantially simultaneously emitted in the monitoring area light signals are generated in each case by a modulated by the spread spectrum with a code output signal, the codes of different light emitter 12 may be generated by quasi-orthogonal Hadamard-Walsh codes or a first light emitter is associated with a first code in the form of an M-sequence pseudo-random code forming a divided into several code segments mother sequence, and at least one further light emitter 12 a further code, which is at least different from the first code, which is generated by a cyclic shift of the code segments of the parent sequence, wherein the different light transmitters 12 associated with different codes again form quasi-orthogonal codes.

The control and / or evaluation device 20 For example, it may be carried out such that n different correlation spectra are evaluated on the receiving side at n at least substantially simultaneously transmitted codes. In this case, the n different correlation spectra can be evaluated in particular linked via the positions of the individual correlation peaks.

The Control and / or evaluation can advantageously also be designed so that at least substantially simultaneously emitted cyclically shifted M-sequences on the receiving side all Codes are identifiable with only one evaluation process. The evaluation can thus in particular by means of a single correlation process be performed with only one correlation spectrum.

at at least substantially simultaneously emitted cyclically shifted M-sequences can be the quasi-orthogonal property of the different Codes based on the autocorrelation properties of the M sequences.

The mother sequence may in particular be divided into at least three code segments. Due to the cyclical shift of the code segments of the parent sequence, at least two and in particular at least three different codes can be generated. The different codes can in particular different light transmitters 12 be associated with a light grid.

The of the light transmitters 12 The light signals emitted at least substantially simultaneously into the monitoring area can in particular be generated in each case by an output signal modulated according to the DSSS (Direct Sequence Spread Spectrum) band spreading method with the relevant code. In doing so, those of the light transmitters 12 be generated at least substantially simultaneously in the monitoring area emitted light signals in each case by at least one mixed with the code in question light pulse.

The of the light receivers 18 In particular, received as a sum signal light signals can be used to identify the a respective light emitter 12 associated output signal with the one respective light emitter 12 assigned code can be acted upon. In this case, the control and / or evaluation 20 medium 20 ' include to identify the a respective light emitter 12 associated output signal to form the autocorrelation function between the received light signals or sum signal and the code in question.

In at least substantially simultaneously emitted quasi-orthogonal Hadamard-Walsh codes, the synchronization of the sensor can be carried out in particular via individual pseudo-random sequences. In to At least substantially simultaneously emitted cyclically shifted M-sequences, the synchronization of the sensor, in particular via the associated from the common light receiver 18 Received sum signal done.

light source 12 and light receiver 18 may be at least partially disposed on opposite sides of the surveillance area. In principle, however, such embodiments are also conceivable, for example, in which light emitter 12 and light receiver 18 on one side of the surveillance area and a retroreflector on the opposite side of the surveillance area. As already mentioned, the light emitters are 12 and light receiver 18 preferably provided in a light grid arrangement. In this case, as already mentioned, in each case a plurality of light transmitters 12 a common light receiver 18 be assigned, so the number of light receivers 18 less than the number of light emitters 12 ,

aim the evaluation strategy is, especially within a light grid to uniquely determine the presence of all simultaneously transmitted codes, so that safe and reliable operation is possible.

Next The code properties also affect the design of the evaluation at. The information contained in the system should be taken into account and be combined in the evaluation strategy in the best possible way.

In the in the 1 reproduced embodiment, by means of the light emitters 12 assigned transmit diodes (here: DAC, Digital Analog Converter) the first binary code C _h [k] with a chip rate of z. B. 4 MHz digital-analog converted.

2 shows an example of a binary code.

The transmitting diodes radiate the analog code signals C _h (t) to the receiving module. The illustrated bandpass characteristic of the code signal results from the bandpass characteristics of the receive module.

3 shows an example of a code signal.

In the optical transmission channel, code signals as well as optical interferers (interference, cf. 1 ) additively to the sum signal R _x (t): R x (t) = C 1 (t) + C 2 (t) + C 3 (t) + I (t)

4 shows an example of a sum signal.

The received sum signal is now sampled at a sampling rate of, for example, 12 MHz and is available as time and value discrete sum signal R _x [k] of the evaluation.

5 shows an example of a discrete sum signal.

The Evaluation takes place via the time code stored in the source code discrete-value reference codes.

6 shows an example of a reference code.

in the The following section deals in detail with the evaluation principle.

In this case, to detect the individual code signals C _h (t), the discrete sum signal R _x [k] can be filtered in three correlation processes with the individual reference codes C _{ref, h} [k]. The correlation is here, for example, as an FIR filter, that is, to be understood as a so-called "matched filter" (FIR = Finite Impulse Response).

7 shows a signal flow diagram of an exemplary evaluation, as well as all other control or evaluation steps in particular by the control and / or evaluation 20 can be carried out.

The evaluation strategy depends on the code type used. This essentially serve the PSR (Peak Sidelobe Ratio) and height and position of the correlation peak as evaluation criteria. The correlation spectrum S _h [m] is given by:

8th shows a correlation spectrum to explain the following terms:

Correlation Peak (Peak of Correlation)

The Reference codes serve as b-coefficients of the individual filters. According to the signal theory is the impulse response of the filter the reference code. Since this is a linear time-invariant System is acting, the system response, with the reference code as Input signal to be a Diracstoss. The Diracstoss is here too referred to as the correlation peak (peak of correlation).

Interference peaks (peak of interference)

By Optical interferers may be in the correlation spectrum Perturbation peaks (peaks of interference) arise.

Correlation noise (sidelobes)

There considered only a finite number of samples of the sum signal and the system is not noise-free (ASIC noise), ent stands in the correlation spectrum so-called correlation noise (sidelobes).

peak

Of the Curve point with maximum ordinate value of the correlation spectrum is called a peak. Here, the abscissa value becomes the peak position (Peak-Pos) and the ordinate value as peak height (peak) designated.

PSR (Peak to Sidelobe Ratio)

The PSR is the ratio of maximum peak to maximum Sidelobes or maximum interference peaks.

Evaluation range

Around computing the PSR will optimize the system response time the evaluation range to a small spectral range reduced. Only in the synchronization is the entire correlation spectrum considered.

Allowed range (Valid Range)

This narrow band in the correlation spectrum is used for synchronization and the signal evaluation. This area can not be on a single Position are reduced because the system requires tolerance for control.

in the Following is the synchronization and control behavior closer considered.

The system has three states that in the in the 9 reproduced state graph are given. The PSR and the height and position of the correlation peak are used for synchronization and control.

The following table shows the state change conditions of the system: ID description 1 PSR ≥ PSR Threshold && Peak ≥ Peak Threshold 2 PSR <PSR Threshold || Peak <Peak Threshold 3 50 times: PSR ≥ PSR Threshold && Peak ≥ Peak Threshold && Peak Pos ε Valid Range 4 100 times: PSR <PSR Threshold || Peak <Peak Threshold

The following are the ones in the 9 and in the preceding table states:

Search State

transmitter and receivers are selectively operated asynchronously. As soon as a peak is detected (condition ID1), the system changes to the state "Sync State".

Sync State

Of the Trigger timer starts the next scan. The start time scanning has an immediate effect on the peak position. The peak position is thus regulated to the center of the correlation spectrum. This process is called synchronization. For example, once the peak position 50 times within the range "Valid Range", the system changes to the state "Secure State".

Secure State

The System will adjust the position of the peak at any time so that the peak position remains within the "Valid Range" range. For example, if the system has no signal over 100 scans more to the scheme available, the system changes to the state "Search State".

in the Below, we will discuss the synchronization principle in more detail.

to Synchronization of the system must clearly detect the signal and be controllable. Are code signal and reference code without displacement τ to each other present, results in a correlation peak centered within the Correlation spectrum.

Are code signal and reference code shifted from each other by τ (cf. 10 ), the correlation peak occurs with the same shift τ in the spectrum (cf. 11 ).

10 So shows the code signal and the reference code with a shift τ, while in the 11 the correlation spectrum is represented with a shift τ of the peak position.

The shift τ is created by sampling too early or too late. The received signal becomes dependent on the displacement τ, as described below: R x [k] = C [k + τ]

The Correlation spectrum is calculated as follows:

For this, the following relationship arises for the peak position: Peak position = τ + N

Under ideal conditions, the following curve is obtained for the peak position as a function of the displacement τ:
The system controls the peak to the center of the correlation spectrum (cf. 12 , Point 1). For error-free synchronization, a steady gradient in the gray area is required in "Sync State".

As soon as the system is synchronized, only the exact position needs to be be readjusted. The regulation in the "Secure State" required for this only in the permitted range (valid range) a steady Course.

12 shows the course of the peak position under ideal conditions.

This History was derived mathematically and does not correspond exactly the actual course of the implemented or simulated System.

at the evaluation is the peak position by means of a signal processing certainly. This may be due to the code signal characteristic result in an unsteady course of the peak position.

• 1. Sprung resultierend aus der Code-Signal-Charakteristik, wobei dieser Sprung auch bei geringer Signaldämpfung auftritt.
• 2. Sporadischer Sprung, der aus der Korrelation von Referenz-Code und ASIC-Rauschen resultiert.

If the system does not detect the correlation peak but peaks other than the peak, the system identifies a discontinuous course of the peak position identified by the system, which can be referred to as a jump in the peak position. The following jump types can be distinguished:

• 1st jump resulting from the code signal characteristic, this jump also occurs at low signal attenuation.
• 2. Sporadic jump resulting from the correlation of reference code and ASIC noise.

• 1. Sync State: Sobald innerhalb des erlaubten Bereichs ein Sprung des Peak-Position auftritt, wird das System nicht in den ”Secure State” gelangen. Treten die Sprünge sporadisch auf, wird der ”Secure State” erreicht. Secure State: Jedoch wird das System auch bei nicht unterbrochenem Schutzfeld, entsprechend dem Sprung, sporadisch auf ”rot” schalten.
• 2. Sync State: Treten Sprünge über den Mittelpunkt des Korrelations-Spektrums auf (von 2.1 nach 2.2 in 13), regelt das System sich um diese Sprünge ein. Da jedoch der erlaubte Bereich nicht erreicht wird, kann das System nicht synchronisieren.
• 3. Sync State: Ergeben sich stetige Verläufe der Peak-Position über den Mittelpunkt des Korrelations-Spektrums bei einer Verschiebung τ ≠ 0, wird sich das System auf diese Verschiebung synchronisieren. Das System wird nicht die gleich hohe Verfügbarkeit gewährleisten wie bei korrekter Synchronisation bei τ = 0, ist jedoch prinzipiell funktionsfähig. Dies sollte dennoch vermieden werden, da bei verschobener Synchronisation nicht die gewünschte Störempfindlichkeit erreicht wird.

If the gradient is not continuous, the following effects can occur during the synchronization in the "Sync State" as well as the rules in the "Secure State":

• 1. Sync State: As soon as a jump in the peak position occurs within the allowed range, the system will not enter the "Secure State". If the jumps occur sporadically, the "secure state" is achieved. Secure State: However, even if the protective field is not interrupted, the system will sporadically switch to "red" according to the jump.
• 2. Sync State: jumps across the center of the correlation spectrum (from 2.1 to 2.2 in 13 ), the system adjusts to these jumps. However, since the allowable range is not reached, the system can not synchronize.
• 3. Sync State: If there are continuous traces of the peak position over the center of the correlation spectrum with a shift τ ≠ 0, the system will synchronize to this shift. The system will not guarantee the same high availability as with the correct synchronization at τ = 0, but is in principle functional. This should nevertheless be avoided, since the desired susceptibility to interference is not achieved with shifted synchronization.

13 shows the course of the peak position with different perturbations.

in the The following is more closely related to the evaluation strategy with a shifted M-sequence (Shifted M-Sequence).

shifted Sequences (Shifted M-Sequences) are an alternative to orthogonal ones Hadamard-Walsh codes. By cyclic shifting of the M sequences can generate "quasi-orthogonal" codes, based on the optimal autocorrelation properties of the M sequences.

The The principle of the shifted M sequence results from the following:

14a shows the typical autocorrelation spectrum of an M-sequence, with the correlation peak being maximal with minimal correlation noise. This correlation peak is also referred to below as the main peak (referred to shifted M sequences) as the main peak.

Code 1 now serves as the "mother sequence" and will appear in 14a shown divided into three code segments S ₁ , S ₂ and S ₃ . Cod 2 is generated by cyclic shifting around a code segment from the code 1.

In a cross-correlation of code 1 and code 2, the spectral energy is distributed over two correlation peaks (cf. 14b ). The positions of the correlation peaks result in the correlation spectrum of the shift of the code segments accordingly. These correlation peaks are also referred to below as side peaks with respect to shifted M sequences.

In the 14 this principle of the shifted M-sequences is reproduced.

As for the synchronization method, synchronization with M-sequences is unaffected a reference code excluded from the outset. The likelihood of synchronizing to a sub-peak is too high. This preliminary consideration is given by the graphs of 15 is occupied, in which the course of the peak position is reproduced with synchronization to a code signal.

Around to avoid erroneous synchronization becomes synchronization the sum signal is used.

16 shows the course of the peak position with a synchronization to the sum signal.

It can now be an extension by linking the correlation spectra be achieved.

So far, the three correlation spectra were evaluated separately. This meant that the PSR, that is, the ratio of peak to maximum sidelobes or maximum spurious peaks, was calculated from S ₁ and checked to see if code 1 was present. Then the PSR of S _{2 was} calculated and checked if code 2 is present, and finally the PSR of S _{3 was} calculated and checked if code 3 is present.

17 now shows the correlation spectra of the three codes.

Since each of the three spectra is calculated from the received sum signal, the displacement τ is the same for all reference codes. This means that the peak position of the main peaks of all correlation spectra must be the same. By the shift τ, the three correlation spectra can be evaluated linked together. The calculation of PSR values is in the 18 shown.

According to 18 So an extended calculation of the PSR for shifted M-sequences is possible.

• – Code-Signal 1 wird von einem Objekt unterbrochen, und zum gleichen Zeitpunkt wird durch einen optischen Störer ein Stör-Peak erzeugt. Annahme: Die Signalform ähnelt der des Code-Signals 1. Hierdurch wird im erlaubten Bereich des Korrelations-Spektrums ein Stör-Peak erzeugt. Dieser Stör-Peak wird nun vom System fälschlicherweise als Haupt-Peak von Code-Signal 1 erkannt.
• – Mit einer separaten Auswertung bleibt das System auf ”grün”, obwohl einer der Strahlen unterbrochen ist, was als äußerst kritisches Ereignis einzustufen ist. ”Grün” gezeichnet den Schaltzustand des Gerätes und bedeutet, dass das Schutzfeld nicht unterbrochen ist und somit der OSSD-Ausgang des Gerätes auf ”high active” geschaltet ist. ”Rot” bedeutet, dass das Schutzfeld unterbrochen ist und somit der OSSD-Ausgang des Gerätes auf ”low” geschaltet ist.

As regards the effects on immunity to interference, the following scenario is considered:

• Code signal 1 is interrupted by an object, and at the same time an interference peak is generated by an optical interferer. Assumption: The signal shape is similar to that of the code signal 1. As a result, an interference peak is generated in the allowed range of the correlation spectrum. This interference peak is now erroneously recognized by the system as the main peak of code signal 1.
• - With a separate evaluation, the system remains "green", even though one of the beams is interrupted, which is classified as a highly critical event. "Green" marks the switching state of the device and means that the protective field is not interrupted and thus the OSSD output of the device is switched to "high active". "Red" means that the protective field is interrupted and thus the OSSD output of the device is switched to "low".

19 illustrates the principle when interference peaks occur.

By the principle of the shifted M sequence is shown by correlation of the Interferer with the code signal 2 the energy of the interfering peak divided into two secondary peaks. This is also the peak position the Stör-peak changed accordingly.

Although the position of the spurious peak (peak Pos ₁ ) in S ₁ is within the allowable range (valid range), it is neither with the position of the spurious peak in S ₂ nor with the position of the correlation peak in S ₂ identical.

The following conclusions can be drawn from the extended evaluation for shifted M sequences:
If one or more of the spurious peaks are erroneously recognized as the main peak, the positions of the main peaks of the three correlation spectra are not identical. This is very important especially in view of the above scenario. In direct comparison, a significant improvement in the FAR (False Acceptance Ratio) can be seen, with the FRR (False Rejection Ratio) only slightly worse.

20 shows a FAR-FRR map of shifted M-sequences with linked evaluation.

The Using shifted M sequences turns out to be very advantageous in terms of signal orthogonality as well the immunity to interference by the linking the correlation spectra.

The Evaluation of the sum signal can be done by two different methods carried out, both methods on the previously explained Principle based on cyclically shifted sequences.

First method:

analysis of the sum signal by means of n correlation processes (n = Number of sequences generated from a parent sequence including the mother sequence; z. For example, if there are three senders to one receiver: n = 3, with five senders to one receiver: n = 6), where the peak positions of the individual correlation spectra are identical. This method has been described in detail previously.

Second method:

Analysis of the sum signal by means of only one correlation process, wherein the reference code is generated by cyclic displacement or extension from the parent sequence as follows:
The mother sequence is divided into n segments (s1, s2, ... sn-1, sn). The individual codes are generated by cyclic shifting as already described. The reference code is now generated by extending the parent sequence by n-1 segments. The segments can be added or appended before or after (synchronous or asynchronous) to the parent sequence. However, it must be ensured that the reference code is a section, with a total of 2 · n - 1 segments, from the following code chain: S1 ... sn S1 S2 ... sn S1 ... sn

through of such a reference code arise in the correlation spectrum n equal correlation peaks. These peaks are caused by the respective transmitted code C1, C2, ... Cn, so that every single code with only one correlation spectrum detected can be.

The in the first method applied correlation of the correlation spectra through the peak positions to minimize the FAR (False Acceptance Ratio) is carried out in this method on the distance of the individual Correlation peak. This is constant at all times and corresponds the displacement of the mother sequence. The checking the peak positions of the n correlation spectra (applied in the first method) is equivalent to the review the distance of the n correlation peaks in the correlation spectrum (the second method).

As already mentioned, the first code can be in the form of a M-sequence, gold sequence or Kasami sequence pseudo-random codes be provided.

10

transmission unit

12

light source

14

optical channel

16

Light-receiving unit

18

light receiver

20

Tax- and / or evaluation unit

20 '

medium

QUOTES INCLUDE IN THE DESCRIPTION

This list The documents listed by the applicant have been automated generated and is solely for better information recorded by the reader. The list is not part of the German Patent or utility model application. The DPMA takes over no liability for any errors or omissions.

Cited patent literature

• - DE 19926214 A1 [0004]

Cited non-patent literature

• - IEC 61496 [0036]

Claims (20)

Opto-electronic sensor with multiple light transmitters ( 12 ) for emitting light signals in a surveillance area, at least one light receiver ( 18 ) for receiving light signals emitted by the light emitters and a control and / or evaluation device ( 20 ) for generating and / or influencing the light signals and for generating an object detection signal in the case of the presence of an object in the surveillance area in dependence on the output signals of the light receiver ( 18 ), characterized in that at least two light emitters ( 12 ) whose light signals emitted at least substantially simultaneously into the monitoring area are respectively generated by an output signal modulated according to the spread spectrum method, the codes of different light transmitters ( 12 ) are generated by quasi-orthogonal Hadamard-Walsh codes or a first light transmitter is associated with a first code in the form of a pseudo-random code comprising an M sequence, gold sequence or Kasami sequence, which comprises a mother sequence divided into several code segments forms, and at least one further light emitter ( 12 ) is associated with a further code different from the first code, which is generated by a cyclic shift of the code segments of the parent sequence, wherein the different light transmitters ( 12 ) again form quasi-orthogonal codes. Optoelectronic sensor according to claim 1, characterized in that the control and / or evaluation device ( 20 ) is executed so that n different correlation spectra are evaluated at n at least substantially simultaneously transmitted codes on the receiving side. Optoelectronic sensor according to claim 2, characterized in that the control and / or evaluation device ( 20 ) is designed so that the n different correlation spectra are evaluated linked via the positions of the individual correlation peaks. Optoelectronic sensor according to one of the preceding claims, characterized in that the control and / or evaluation device ( 20 ) is designed so that at least substantially simultaneously emitted cyclically shifted M-sequences, gold sequences or Kasami sequences on the receiving side all codes are identifiable with only one evaluation process. Optoelectronic sensor according to one of the preceding Claims, characterized in that at least in Essentially simultaneously emitted cyclically shifted M-sequences, Gold sequences or Kasami sequences the quasi-orthogonal property different codes on the autocorrelation properties M sequences, gold sequences or Kasami sequences based. Optoelectronic sensor according to one of the preceding Claims, characterized in that the mother sequence in at least two, in particular in at least three and preferably is divided into two or three to six code segments. Optoelectronic sensor according to one of the preceding Claims, characterized in that by the cyclic Shift the code segments of the mother sequence at least two, in particular three and preferably two or three to six different Codes are generated. Optoelectronic sensor according to one of the preceding claims, characterized in that the different codes are different light emitters ( 12 ) are assigned to a light grid. Optoelectronic sensor according to one of the preceding claims, characterized in that the light emitters ( 12 ) are generated at least substantially simultaneously in the monitoring area emitted light signals in each case by a DSSS (Direct Sequence Spread Spectrum) -Bandspreizverfahren modulated with the code in question output signal. Optoelectronic sensor according to one of the preceding claims, characterized in that the light emitters ( 12 ) are generated at least substantially simultaneously in the monitoring area emitted light signals in each case by at least one mixed with the code in question light pulse. Optoelectronic sensor according to one of the preceding claims, characterized in that that of the light receiver ( 18 ) in particular as a sum signal received light signals for identifying the a respective light emitter ( 12 ) associated with the output signal Light transmitter ( 12 ) associated code are applied. Optoelectronic sensor according to claim 9, characterized in that the control and / or evaluation device ( 20 ) Medium ( 20 ' ) in order to identify the respective light emitter ( 12 ) to form the autocorrelation function between the received light signals and the respective code. Optoelectronic sensor according to one of the preceding claims, characterized in that in each case at least two light emitters ( 12 ) a common light receiver ( 18 ) assigned. Optoelectronic sensor according to one of the preceding claims, characterized in that in each case at least two, in particular three and preferably two or three to six light transmitters ( 12 ) a common light receiver ( 18 ) assigned. Optoelectronic sensor according to one of the preceding Claims, characterized in that at least essentially simultaneously emitted quasi-orthogonal Hadamard-Walsh codes the synchronization of the sensor via individual pseudo-random sequences he follows. Optoelectronic sensor according to one of the preceding claims, characterized in that at least substantially simultaneously emitted cyclically shifted M-sequences, gold sequences or Kasami sequences, the synchronization of the sensor via the associated from the common light receiver ( 18 ) received sum signal takes place. Optoelectronic sensor according to one of the preceding claims, characterized in that the light transmitter ( 12 ) and light receiver ( 18 ) are arranged at least partially on opposite sides of the surveillance area. Optoelectronic sensor according to one of the preceding claims, characterized in that the light transmitter ( 12 ) and light receiver ( 18 ) on one side of the surveillance area and a retroreflector on the opposite side of the surveillance area. Optoelectronic sensor according to one of the preceding claims, characterized in that the light emitters ( 12 ) and light receiver ( 18 ) are provided in a light grid arrangement. Optoelectronic sensor according to one of the preceding claims, characterized in that in each case a plurality of light transmitters ( 12 ) a common light receiver ( 18 ) and correspondingly the number of light receivers ( 18 ) is less than the number of light emitters ( 12 ).